This project is comprised of two components that address related issues, access to VA health care for lower eligibility priority veterans and optimal staffing for VA health care delivery.
The objective for the access component is to describe trends in eligibility status conversions of lower eligibility priority veterans to higher eligibility priorities for receiving VA health care. The objectives for the staffing component of the project were to: 1) develop optimal staffing levels for 45 employment categories of VA FTEE, stratified by Veterans Integrated Service Network (VISN) and VA medical center (VAMC) and; 2) measure changes in VA efficiency (as measured by cost) since the Department of Veteran Affairs reorganization into Veteran Integrated Service Networks.
The access component of the study employed a retrospective cohort design and multivariate logistic regression analyses. The optimal staffing component employed stochastic frontier, simultaneous equation and panel data analyses using data from the national VA database housed at the Austin Automation Center.
Lower priority (LP) veterans accounted for 4.6% of VA workload in FY95 and 21.6% of the VA workload in FY01, a 563% increase. Five-year observation of a cohort of LP patients, revealed that 55.5% of LP VA health care users continued to use VA services and 52% of those users eventually converted to high priority (HP) eligibility status. Therefore, VA means test conversion from LP to HP status has substantial impact on VA workload and resources. Demographic factors contributed to this eligibility-status change - for example, Minority and Persian Gulf War veterans had relatively high odds ratios of changing their eligibility status to HP Groups. Also eligibility status conversion displayed a geographic pattern where LP VA users living in "snow belt" regions are more likely to stay in LP groups and LP VA users living in "sun belt" regions are more likely to convert to HP eligibility status.
Staffing levels, managerial and operational efficiency (as measured by cost) was assessed in the second component of the study. A wide range of patient variables and VAMC characteristics were needed to conduct the assessment. In these analyses every patient using VA from FY95 to FY02 was risk-adjusted and their demographic information such as income and percent of service-connected disability rating was used. Further, we collected data for total cost and FTEEs in 45 employment categories by each VAMC. We also collected VAMC characteristics such as amount of snowfall, footage and age of VAMC buildings, and number of affiliated Community Based Outpatient Clinics. Some variables that are commonly believed to be causes of higher cost, such as parking lot and campus size did not independently predict costs after controlling for other variables in the models. The panel data analysis revealed that VA had gained 14% in managerial or operational efficiency between FY95 and FY02. However, room for additional improvement in efficiency (e.g. administrative personnel) was found. By using the stochastic frontier, simultaneous equations and panel data techniques, a framework that estimates the optimal staffing levels for each VAMC and identifies VAMCs that were over or understaffed for the study period was developed and implemented.
The study has important policy implications because it examines patterns of eligibility for access in LP veterans, the fastest growing component of VA health care users. The optimal staffing component of the study provides VA with a method to evaluate staffing levels.
External Links for this Project
None at this time.